I am using a cromwell server running on our kubernetes cluster on GCP. I see, in the execution bucket, copies of my (large) input files. Why might I be seeing these extraneous copies?
Hmm, that is a little surprising - are those files are being assigned as outputs to any of your tasks?
Ok... I'll make this a ticket in the Cromwell repo. One other question - do you happen to have an output glob that might be gathering up the input files as well as their intended output?
Hi, following up on this, and assuming that the answer to Chris' question above is "no", could you share a workflow that produces this behavior along with the Cromwell configuration ?